Logo des Repositoriums

Auflistung nach Schlagwort "ground truth"

1 - 2 von 2
  • Konferenzbeitrag
    A new Attack Composition for Network Security
    (10. DFN-Forum Kommunikationstechnologien, 2017) Beer, Frank; Hofer, Tim; Karimi, David; Bühler, Ulrich
    As the current cyber threat landscape is becorning more depressing, sophisticated intru­sion detection systems must evolve to protect network infrastructures efficiently. Building such a detector is highly data-driven and requires quality datasets to evaluate different phases in both the development and deployment process. However, finding publicly available captures with a ground truth is challenging. Most existing datasets focus on very specific subjects such as botnet, flooding, or brute-force traffic rather than providing a broad arsenal of different attack vectors threatening today's networks. This work addresses this gap by introducing a new attack composition compris­ing a multitude of classic as well as state-of-the-art attacks. The dataset embrace rich and untreated packet traces including payload, collected log events, and a detailed ground truth. lntitial results re­veal the proposed captures complement existing traces and provide a sound base for various rnining applications in the field of network security research.
  • Konferenzbeitrag
    Utility-based performance evaluation of biometric sample quality assessment algorithms
    (BIOSIG 2022, 2022) Olaf Henniger, Biying Fu and Cong Chen
    The quality score of a biometric sample is expected to predict the sample’s utility, but a universally valid definition of utility is missing. A harmonized definition of utility would be useful to facilitate the comparison of biometric sample quality assessment algorithms. This paper generalizes the utility of a biometric sample as normalized difference between the means of non-mated and mated comparison scores with respect to this sample. Using a face image data set, we show that discarding samples with low utility scores determined in this way results in a rapidly declining false non-match rate. The obtained utility scores can be used as ground-truth utility labels for training biometric sample quality assessment algorithms and for summarizing their prediction performance in a single plot and in a single figure of merit based on the proposed utility score definition.